CN106558040B - Character image treating method and apparatus - Google Patents
Character image treating method and apparatus Download PDFInfo
- Publication number
- CN106558040B CN106558040B CN201510613580.6A CN201510613580A CN106558040B CN 106558040 B CN106558040 B CN 106558040B CN 201510613580 A CN201510613580 A CN 201510613580A CN 106558040 B CN106558040 B CN 106558040B
- Authority
- CN
- China
- Prior art keywords
- processed
- human body
- pixel
- body area
- width
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 31
- 238000012545 processing Methods 0.000 claims abstract description 75
- 238000001514 detection method Methods 0.000 claims description 22
- 230000009466 transformation Effects 0.000 claims description 22
- 238000003672 processing method Methods 0.000 claims description 18
- 238000003708 edge detection Methods 0.000 claims description 10
- 238000004590 computer program Methods 0.000 claims description 3
- 230000008569 process Effects 0.000 abstract description 10
- 230000008439 repair process Effects 0.000 abstract description 5
- 210000000746 body region Anatomy 0.000 description 33
- 238000010586 diagram Methods 0.000 description 13
- 230000006870 function Effects 0.000 description 7
- 210000004705 lumbosacral region Anatomy 0.000 description 6
- 235000002918 Fraxinus excelsior Nutrition 0.000 description 4
- 239000002956 ash Substances 0.000 description 4
- 238000006243 chemical reaction Methods 0.000 description 4
- 239000000284 extract Substances 0.000 description 3
- 241000208340 Araliaceae Species 0.000 description 2
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 description 2
- 235000003140 Panax quinquefolius Nutrition 0.000 description 2
- 238000004422 calculation algorithm Methods 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 235000008434 ginseng Nutrition 0.000 description 2
- 239000003973 paint Substances 0.000 description 2
- 230000008859 change Effects 0.000 description 1
- 230000005611 electricity Effects 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T3/00—Geometric image transformations in the plane of the image
- G06T3/40—Scaling of whole images or parts thereof, e.g. expanding or contracting
- G06T3/4084—Scaling of whole images or parts thereof, e.g. expanding or contracting in the transform domain, e.g. fast Fourier transform [FFT] domain scaling
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The present invention relates to a kind of character image treating method and apparatus.It the described method comprises the following steps: obtaining image；The face location in described image is detected, and marks the parameter of the face；According to the parameter of human body area to be processed in the parameter estimation described image of the face；Human body area to be processed is determined according to the parameter of human body area to be processed；Obtain the width at human body position to be processed in human body area to be processed；Processing is zoomed in and out to the pixel value of each pixel in the width at human body position to be processed in human body area to be processed, the pixel value after obtaining each pixel scaling.Above-mentioned character image treating method and apparatus realizes the processing of the pixel to the position to be processed of human body in character image automatically, and do not need to spend high cost study profession repairs figure software, automatically processes to personage in picture, high-efficient.
Description
Technical field
The present invention relates to field of image processings, more particularly to a kind of character image treating method and apparatus.
Background technique
Currently, popularizing with electronic products such as digital camera, shooting mobile phone, cameras, more and more users pass through
Electronic product with shooting function carries out self-timer or shoots image for other people.But it due to light, apparatus for making a video recording, personal appearance, claps
The reasons such as angle, shooting posture, lens distortion are taken the photograph, the image effect after some shootings is poor, especially the bodily form, often not to the utmost such as people
Meaning, especially some users always think that the bodily form is too fat, influence the overall effect of image.For this purpose, passing through the personage of some professions
It is handled using softwares such as photoshop, the bodily form is reduced, makes personage very thin, it is beautiful generous.However, using photoshop etc.
Professional software learning cost is high, and trouble is manually operated, and treatment effeciency is low.
Summary of the invention
Based on this, it is necessary to aiming at the problem that traditional personage handles manual operation trouble and low efficiency, provide a kind of people
Object image processing method can automatically process image, improve treatment effeciency.
In addition, there is a need to provide a kind of character image processing unit, image can be automatically processed, improves treatment effeciency.
A kind of character image processing method, comprising the following steps:
Obtain image；
The face location in described image is detected, and marks the parameter of the face；
According to the parameter of human body area to be processed in the parameter estimation described image of the face；
Human body area to be processed is determined according to the parameter of human body area to be processed；
Obtain the width at human body position to be processed in human body area to be processed；
To the pixel value of each pixel in the width at human body position to be processed in human body area to be processed into
Row scaling processing, the pixel value after obtaining each pixel scaling.
A kind of character image processing unit, comprising:
Image collection module, for obtaining image；
Mark module is detected, for detecting the face location in described image, and marks the parameter of the face；
Estimation block, the ginseng for human body area to be processed in the parameter estimation described image according to the face
Number；
Area determination module, for determining human body position area to be processed according to the parameter of human body area to be processed
Domain；
Width detection module, for obtaining the width at human body position to be processed in human body area to be processed；
Zoom module, for each pixel in the width to human body position to be processed in human body area to be processed
The pixel value of point zooms in and out processing, the pixel value after obtaining each pixel scaling.
Above-mentioned character image treating method and apparatus, after obtaining image, face location in detection image marks the ginseng of face
Number is determined according to the parameter of face parameter estimation human body area to be processed according to the parameter of human body area to be processed
Human body area to be processed obtains the width at human body position to be processed in human body area to be processed, to pixel therein
The pixel of point zooms in and out processing, and the image after being scaled realizes automatically to the position to be processed of human body in character image
The processing of pixel, do not need to spend high cost study profession repairs figure software, automatically processes to personage in picture, efficiency
It is high.
Detailed description of the invention
Fig. 1 is the schematic diagram of internal structure of the terminal with shooting function in one embodiment；
Fig. 2 is the flow chart of character image processing method in one embodiment；
Fig. 3 is the flow chart of character image processing method in another embodiment；
Fig. 4 is the schematic diagram of face label；
Fig. 5 is the schematic diagram of the position of each parameter of people's body body region；
Fig. 6 is the schematic diagram of the position of each parameter of human body lumbar region；
Fig. 7 is the structural block diagram of character image processing unit in one embodiment；
Fig. 8 is the structural block diagram of character image processing unit in another embodiment.
Specific embodiment
In order to make the objectives, technical solutions, and advantages of the present invention clearer, with reference to the accompanying drawings and embodiments, right
The present invention is further elaborated.It should be appreciated that the specific embodiments described herein are merely illustrative of the present invention, and
It is not used in the restriction present invention.
Fig. 1 is the schematic diagram of internal structure of the terminal with shooting function in one embodiment.As shown in Figure 1, the terminal packet
Include processor, storage medium, memory, display screen and the input unit connected by system bus.Wherein, the storage medium of terminal
It is stored with operating system, further includes a kind of character image processing unit, the character image processing unit is for realizing a kind of personage
Image processing method.The processor supports the operation of entire terminal for providing calculating and control ability.It is saved as in terminal
The operation of character image processing unit in storage medium provides environment.The display screen of terminal can be liquid crystal display or electricity
Sub- ink display screen etc., input unit can be the touch layer covered on display screen, be also possible to be arranged in terminal enclosure by
Key, trace ball or Trackpad are also possible to external keyboard, Trackpad or mouse etc..The terminal can be mobile phone, tablet computer
Or personal digital assistant etc..It will be understood by those skilled in the art that structure shown in Fig. 1, only and application scheme
The block diagram of relevant part-structure does not constitute the restriction for the terminal being applied thereon to application scheme, specific terminal
It may include perhaps combining certain components or with different component layouts than more or fewer components as shown in the figure.
Fig. 2 is the flow chart of character image processing method in one embodiment.As shown in Fig. 2, a kind of character image processing
Method, comprising the following steps:
Step 202, image is obtained.
Specifically, image can acquire equipment acquisition by camera etc..The image of acquisition can be encoded, coding
Format can be BMP (Bit Map, bitmap), JPEG (Joint Photographic Expert Group), GIF (Graphics
Interchange Format, graph transformation format), TGA (Tagged Graphics, marked figure), EXIF
(Exchangeable Image File Format, tradable image file format), PNG (Portable Network
Graphics, portable network figure) etc..
It, can be by the image decoding after coding at original image information if what is obtained is the image after coding when obtaining image
Stream, such as the image data of RGB (Red Green Blue) format, is then handled decoded image.
Step 204, the face location in the image is detected, and marks the parameter of the face.
Specifically, detection face location can export the coordinate of face frame automatically, determine face according to the coordinate of face frame
Region, the parameter of face include the center point coordinate of face, face width and face height etc..Mark the center point coordinate of face
(fx, fy), face width and height are respectively (wf, hf)。
Step 206, according to the parameter of human body area to be processed in the parameter estimation of the face image.
Specifically, step 206 includes: according to the center point coordinate of face, width and height estimation human body position to be processed
Lower line coordinates, upper line coordinates, width and the center point coordinate in region etc..
Human body position to be processed can be body or waist or face or leg etc..Human body area to be processed, which refers to, includes
One image range at human body position to be processed.
Step 208, human body area to be processed is determined according to the parameter of human body area to be processed.
Specifically, automatically according to the lower line coordinates of the human body area to be processed in image, upper line coordinates, width and in
Heart point coordinate can determine the human body area to be processed in image.
Step 210, the width at human body position to be processed in human body area to be processed is obtained.
In the present embodiment, the edge detection of vertical direction can be carried out to human body area to be processed automatically, obtains human body
The both sides of the edge line at position to be processed, obtain both sides of the edge line the distance between obtain the width at human body position to be processed.
Specifically, the edge detection of vertical direction is carried out to human body area to be processed, extracts the longest edge in two sides
As the both sides of the edge line of human body area to be processed, calculates the distance between both sides of the edge line and obtain human body position to be processed
Width.
The width at human body position to be processed refers to the developed width at human body position to be processed in image.
Step 212, to the pixel of each pixel in the width at human body position to be processed in human body area to be processed
Value zooms in and out processing, the pixel value after obtaining each pixel scaling.
Specifically, scaling processing can be zoomed in and out according to scaling, or according to the zoom factor function pre-established
Zoom in and out processing.
In one embodiment, to each pixel in the width at human body position to be processed in human body area to be processed
The step of pixel value of point zooms in and out processing, obtains the pixel value after each pixel scales include:
Each pixel obtained in human body area to be processed in the width at human body position to be processed is to be processed to human body
Zoom factor corresponding to the horizontal distance of the central point of area and each pixel；
If the horizontal distance of certain pixel is less than the head width of presupposition multiple and is located at human body area to be processed
Central point the first side, then according to the central point of human body area to be processed and the corresponding zoom factor of certain pixel
First object pixel is obtained, the pixel of predetermined quantity around the first object pixel is chosen, by the picture of the predetermined quantity
The pixel value of vegetarian refreshments carries out interpolation and obtains the pixel value of interpolation, using the pixel value of the interpolation as the new of certain pixel
Pixel value；
If the horizontal distance of certain pixel is less than the head width of presupposition multiple and is located at human body area to be processed
Central point the second side, then according to the central point of human body area to be processed and the corresponding zoom factor of certain pixel
The second target pixel points are obtained, the pixel of predetermined quantity around second target pixel points is chosen, by the picture of the predetermined quantity
The pixel value of vegetarian refreshments carries out interpolation and obtains the pixel value of interpolation, using the pixel value of the interpolation as the new of certain pixel
Pixel value.
Wherein, zoom factor can be calculated according to smooth transformation formula, smooth transformation formula such as formula (1) (2) (3):
Wherein, r is the horizontal distance of the central point of pixel and human body area to be processed；D is the width on head, i.e.,
wf；A and b is constant, and the value of a and b can be determined as needed.In the present embodiment, a 1, b 0.8.Smooth transformation formula
It is according to constantly test, it is ensured that edge variation is small, and intermediate conversion is big.θ (r), λ (r) are intermediate variable, and f (r) is contracting
Put coefficient.
Each pixel (x, y) in human body area to be processed in the width at human body position to be processed, if its with
Central point (the x of human body area to be processed0, y0) horizontal distance be r, if r be less than 2d, and be located at human body portion to be processed
The first side (such as left side) of the central point in position region, then obtaining first object pixel is (x0,y0- f (r)), if r is less than 2d, and
Positioned at the second side (such as the right) of the central point of human body area to be processed, then obtaining the second target pixel points is (x0,y0+f
(r)).Wherein, the left side and the right when left and right refers to that image is checked, when personage is in facing to viewer position.
Predetermined quantity can be chosen as needed, such as four pixels, five pixels, six pixels etc..If choosing
Four pixels, then the pixel value obtained to the pixel value of four pixels using bilinear interpolation algorithm operation is as the first mesh
Mark the pixel value of pixel or the second target pixel points.Bilinear interpolation first can carry out linear interpolation in X-direction, then in the Y direction
Upper carry out linear interpolation, or linear interpolation is first carried out in the Y direction, then carries out linear interpolation in the X direction.
When certain pixel is located at the first side of the central point of human body area to be processed, according to human body portion to be processed
The central point in position region and the corresponding zoom factor of certain pixel the step of obtaining first object pixel include:
The abscissa and ordinate for obtaining the central point of human body area to be processed, by human body position area to be processed
Abscissa of the abscissa of the central point in domain as first object pixel, by the central point of human body area to be processed
Ordinate subtracts ordinate of the difference of the corresponding zoom factor of certain pixel as first object pixel, according to first mesh
The abscissa and ordinate for marking pixel determine first object pixel.
It is to be processed according to the human body when certain pixel is located at the second side of the central point of human body area to be processed
The step of central point of area and the corresponding zoom factor of certain pixel obtain the second target pixel points include:
The abscissa and ordinate for obtaining the central point of human body area to be processed, by human body position area to be processed
Abscissa of the abscissa of the central point in domain as the second target pixel points, by the central point of human body area to be processed
Ordinate adds ordinate of the sum of the corresponding zoom factor of certain pixel as the second target pixel points, according to second mesh
The abscissa and ordinate for marking pixel determine second target pixel points.
Above-mentioned character image processing method, after obtaining image, face location in detection image marks the parameter of face, root
According to the parameter of face parameter estimation human body area to be processed, determine that human body waits for according to the parameter of human body area to be processed
Treatment site region obtains the width at human body position to be processed in human body area to be processed, to the picture of pixel therein
Element zooms in and out processing, and the image after being scaled realizes automatically to the pixel at the position to be processed of human body in character image
Processing, do not need to spend high cost study profession repairs figure software, automatically processes to personage in picture, high-efficient.
Fig. 3 is the flow chart of character image processing method in another embodiment.Character image processing method and figure in Fig. 3
2 difference is, increases and judges whether the width at human body position to be processed is more than threshold value, is more than then to be adjusted, if not surpassing
It crosses, does not then adjust, and histogram equalization processing is carried out to the image after scaling processing.As shown in figure 3, a kind of character image
Processing method, comprising the following steps:
Step 302, image is obtained.
Specifically, image can acquire equipment acquisition by camera etc..The image of acquisition can be encoded, coding
Format can be BMP, JPEG, GIF, TGA, EXIF, PNG etc..
It, can be by the image decoding after coding at original image information if what is obtained is the image after coding when obtaining image
Stream, such as the image data of rgb format, is then handled decoded image.
Step 304, the face location in the image is detected, and marks the parameter of the face.
Specifically, detection face location can export the coordinate of face frame automatically, determine face according to the coordinate of face frame
Region, the parameter of face include the center point coordinate of face, face width and face height etc..Mark the center point coordinate of face
(fx, fy), face width and height are (wf, hf)。
Step 306, according to the parameter of human body area to be processed in the parameter estimation of the face image.
Specifically, step 306 includes: according to the center point coordinate of face, width and height estimation human body position to be processed
Lower line coordinates, upper line coordinates, width and the center point coordinate in region etc..
Human body position to be processed can be body or waist or face or leg etc..Human body area to be processed, which refers to, includes
One image range at human body position to be processed.
Step 308, human body area to be processed is determined according to the parameter of human body area to be processed.
Specifically, automatically according to the lower line coordinates of the human body area to be processed in image, upper line coordinates, width and in
Heart point coordinate can determine the human body area to be processed in image.
Step 310, the width at human body position to be processed in human body area to be processed is obtained.
In the present embodiment, the edge detection of vertical direction can be carried out to human body area to be processed automatically, obtains human body
The both sides of the edge line at position to be processed, obtain both sides of the edge line the distance between obtain the width at human body position to be processed.
Specifically, the edge detection of vertical direction is carried out to human body area to be processed, extracts the longest edge in two sides
As the both sides of the edge line of human body area to be processed, calculates the distance between both sides of the edge line and obtain human body position to be processed
Width.
The width at human body position to be processed refers to the developed width at human body position to be processed in image.
Step 312, judge whether the width at human body position to be processed in human body area to be processed is greater than threshold value, if
It is, step 314, if it is not, thening follow the steps 318.
Step 314, to the pixel of each pixel in the width at human body position to be processed in human body area to be processed
Value zooms in and out processing, the pixel value after obtaining each pixel scaling.
Specifically, scaling processing can be zoomed in and out according to scaling, or according to the zoom factor function pre-established
Zoom in and out processing.
Each pixel obtained in human body area to be processed in the width at human body position to be processed is to be processed to human body
Zoom factor corresponding to the horizontal distance of the central point of area and each pixel；
If the horizontal distance of certain pixel is less than the head width of presupposition multiple and is located at human body area to be processed
Central point the first side, then according to the central point of human body area to be processed and the corresponding zoom factor of certain pixel
First object pixel is obtained, the pixel of predetermined quantity around the first object pixel is chosen, by the picture of the predetermined quantity
The pixel value of vegetarian refreshments carries out interpolation and obtains the pixel value of interpolation, using the pixel value of the interpolation as the new of certain pixel
Pixel value；
If the horizontal distance of certain pixel is less than the head width of presupposition multiple and is located at human body area to be processed
Central point the second side, then according to the central point of human body area to be processed and the corresponding zoom factor of certain pixel
The second target pixel points are obtained, the pixel of predetermined quantity around second target pixel points is chosen, by the picture of the predetermined quantity
The pixel value of vegetarian refreshments carries out interpolation and obtains the pixel value of interpolation, using the pixel value of the interpolation as the new of certain pixel
Pixel value.
Step 316, without processing.
Step 318, histogram equalization processing is carried out to the image after scaling.
Specifically, the contrast of image can be improved in histogram equalization processing, and it is uneven to reduce transformation bring color.
Histogram equalization processing is to become the grey level histogram of original image in whole ashes from some gray scale interval for comparing concentration
Being uniformly distributed in degree range.
In addition, the method may also include that the image after showing histogram equalization processing after step 318.Specifically
Ground shows the image after histogram equalization processing on the interface of terminal.
Above-mentioned character image processing method, after obtaining image, face location in detection image marks the parameter of face, root
According to the parameter of face parameter estimation human body area to be processed, determine that human body waits for according to the parameter of human body area to be processed
Treatment site region obtains the width at human body position to be processed in human body area to be processed, to the picture of pixel therein
Element zooms in and out processing, and the image after being scaled realizes automatically to the pixel at the position to be processed of human body in character image
Processing, do not need to spend high cost study profession repairs figure software, automatically processes to personage in picture, high-efficient；Sentence
Disconnected width can reduce the data volume of processing；The contrast of image can be improved in histogram equalization processing, reduces transformation bring
Color is uneven.
In other embodiments, step 318 can also be omitted.
Character image processing method of the invention is described below in detail and is applied to processing character physical region, detailed process
Include:
(a1) image is obtained, by image decoding at rgb format；
(a2) face location in detection image, marks the parameter of face, and face parameter includes that the center point coordinate of face is
(fx, fy), face width and height be respectively (wf, hf)。
Fig. 4 is the schematic diagram of face label.As shown in figure 4, the center point coordinate of face is (fx, fy), face width and height
Degree is respectively (wf, hf)。
(a3) according to the parameter in the parameter estimation human body region of face, human body region parameter includes human body
Lower line coordinates, upper line coordinates, width and the center point coordinate in region.
Specifically, the body paint rule for finding and proposing according to Leonardo da Vinci: the ratio of standardized human body is that head is height
1/8, shoulder breadth is the 1/4 of height, with navel be, upper lower part of the body ratio be 5:8, meet " golden section " law, with positive dough figurine
As for, the parameter calculation formula in human body region is as follows:
Human body total height: high=8*hf
Body region it is offline are as follows: hby=fy-7.5*hf
Body region it is online are as follows: hty=hby+8hf=fy+0.5hf
The width of body region is denoted as head width 3 times: wc=3*wf
The center point coordinate of body region are as follows: (fx, fy-3.5hf)
Fig. 5 is the schematic diagram of the position of each parameter of people's body body region.As shown in figure 5, the offline of body region is people
Body starting point coordinate hby, the online of body region is hty, the width of body region is wc。
(a4) human body region is determined according to the parameter in human body region.
Specifically, human body region is between offline on body region, with body region centre coordinate (fx, fy-
3.5hf) centered on, width wcRegion.
(a5) the width w of body is detected1。
Specifically, the edge detection that vertical direction is carried out in human body region, obtains the both sides of the edge of human body
Line, obtain both sides of the edge line the distance between obtain the width w of human body1。
(a6) judge whether the width of body is greater than body threshold value, if so, needing to adjust, that is, converted, if it is not, then
It is not required to adjust.
Specifically, body threshold value can be 2 times of head width 2*wf。
(a7) zoom factor is obtained.
Zoom factor can be calculated according to smooth transformation formula, smooth transformation formula such as formula (1) (2) (3):
Wherein, r is the horizontal distance of the central point in pixel and human body region；D is the width on head, i.e. wf；A and
B is constant, and the value of a and b can be determined as needed.In the present embodiment, a 1, b 0.8.Smooth transformation formula is basis
Constantly test, it is ensured that edge variation is small, and intermediate conversion is big.θ (r), λ (r) are intermediate variable, and f (r) is scaling system
Number.
(a8) pixel in the width of human body is converted.
Each pixel (x, y) in human body region in the width of body, if its in human body region
Heart point (fx, fy-3.5hf) horizontal distance be r, if r be less than 2d, and be located at human body body region central point the first side
(such as left side), then obtaining first object pixel is (fx, fy-3.5hf- f (r)), choose first object pixel (fx, fy-
3.5hf- f (r)) around four pixels bilinear interpolation after pixel be the pixel pixel value；If r is less than 2d, and
Positioned at the second side (such as the right) of the central point in human body region, then obtaining the second target pixel points is (fx, fy-3.5hf+f
(r)) first object pixel (f, is chosenx, fy-3.5hf+ f (r)) around four pixels bilinear interpolation after pixel be
The pixel value of the pixel.
(a9) histogram equalization processing is carried out to by transformed image.
Specifically, the contrast of image can be improved in histogram equalization processing, and it is uneven to reduce transformation bring color.
Histogram equalization processing is to become the grey level histogram of original image in whole ashes from some gray scale interval for comparing concentration
Being uniformly distributed in degree range.
Character image processing method of the invention is described below in detail and is applied to processing personage lumbar region, detailed process
Include:
(b1) image is obtained, by image decoding at rgb format；
(b2) face location in detection image, marks the parameter of face, and face parameter includes that the center point coordinate of face is
(fx, fy), face width and height be respectively (wf, hf)。
Join shown in Fig. 4 again, the center point coordinate of face is (fx, fy), face width and height be respectively (wf, hf)。
(b3) according to the parameter in the parameter estimation human body waist region of face, human body waist region parameter includes human body waist
Lower line coordinates, upper line coordinates, width and the center point coordinate in region.
Specifically, the body paint rule for finding and proposing according to Leonardo da Vinci: the ratio of standardized human body is that head is height
1/8, shoulder breadth is the 1/4 of height, with navel be, upper lower part of the body ratio be 5:8, meet " golden section " law, with positive dough figurine
As for, the parameter calculation formula in human body region is as follows:
Human body total height: high=8*hf
Body region it is offline are as follows: hby=fy-4.5*hf
Body region it is online are as follows: hty=hby+2hf=fy-2.5hf
The width of body region is denoted as head width 3 times: we=2*wf
The center point coordinate of body region are as follows: (fx, fy-3.5hf)
Fig. 6 is the schematic diagram of the position of each parameter of human body lumbar region.As shown in fig. 6, the offline of body region is people
Body starting point coordinate hby, the online of lumbar region is hty, the width of lumbar region is we。
(b4) human body waist region is determined according to the parameter in human body waist region.
Specifically, human body region is between offline on body region, with body region centre coordinate (fx, fy-
3.5hf) centered on, width weRegion.
(b5) the width w of waist is detected2。
Specifically, the edge detection that vertical direction is carried out in human body waist region, obtains the both sides of the edge of human body waist
Line, obtain both sides of the edge line the distance between obtain the width w of human body waist2。
(b6) judge whether the width of waist is greater than waist threshold value, if so, needing to adjust, that is, converted, if it is not, then
It is not required to adjust.
Specifically, threshold value can be 2 times of head width 2*wf。
(b7) zoom factor is obtained.
Zoom factor can be calculated according to smooth transformation formula, smooth transformation formula such as formula (1) (2) (3):
Wherein, r is the horizontal distance of pixel and human body waist central point；D is the width on head, i.e. wf；A and b is normal
The value of number, a and b can be determined as needed.In the present embodiment, a 1, b 0.8.Smooth transformation formula is according to constantly examination
It tests, it is ensured that edge variation is small, and intermediate conversion is big.θ (r), λ (r) are intermediate variable, and f (r) is zoom factor.
(b8) pixel in the width of human body waist is converted.
Each pixel (x, y) in human body region in the width of waist, if its in human body waist region
Heart point (fx, fy-3.5hf) horizontal distance be r, if r be less than 2d, and be located at human body lumbar region central point the first side
(such as left side), then obtaining first object pixel is (fx, fy-3.5hf- f (r)), choose first object pixel (fx, fy-
3.5hf- f (r)) around four pixels bilinear interpolation after pixel be the pixel pixel value；If r is less than 2d, and
Positioned at the second side (such as the right) of the central point in human body waist region, then obtaining the second target pixel points is (fx, fy-3.5hf+f
(r)) first object pixel (f, is chosenx, fy-3.5hf+ f (r)) around four pixels bilinear interpolation after pixel be
The pixel value of the pixel.
(b9) histogram equalization processing is carried out to by transformed image.
Specifically, the contrast of image can be improved in histogram equalization processing, and it is uneven to reduce transformation bring color.
Histogram equalization processing is to become the grey level histogram of original image in whole ashes from some gray scale interval for comparing concentration
Being uniformly distributed in degree range.
Fig. 7 is the structural block diagram of character image processing unit in one embodiment.Character image processing unit pair in Fig. 7
The virtual functions framework constructed by the character image processing method of Fig. 2, the not detailed place of description, which please refers in method, retouches
State, the functional module in Fig. 7 be not limited to the present embodiment described in division mode.As shown in fig. 7, a kind of character image processing
Device, including image collection module 710, detection mark module 720, estimation block 730, area determination module 740, width detection
Module 750 and Zoom module 760.Wherein:
Image collection module 710 is for obtaining image.
It specifically,, can be by the image decoding after coding at original graph if what is obtained is the image after coding when obtaining image
Then piece information flow is handled decoded image such as the image data of RGB (Red Green Blue) format.
Detection mark module 720 is used to detect the face location in the image, and marks the parameter of the face.
Specifically, detection face location can export the coordinate of face frame automatically, determine face according to the coordinate of face frame
Region, the parameter of face include the center point coordinate of face, face width and face height etc..Mark the center point coordinate of face
(fx, fy), face width and height are respectively (wf, hf)。
Detection mark module 720 is also used to mark the center point coordinate of the face, the width of face and height.
Estimation block 730 is used for the parameter according to human body area to be processed in the parameter estimation of the face image.
Specifically, estimation block 730 is according to the center point coordinate of face, width and height estimation human body position area to be processed
Lower line coordinates, upper line coordinates, width and the center point coordinate in domain etc..Human body position to be processed can for body or waist or face or
Leg etc..Human body area to be processed refers to an image range comprising human body position to be processed.
Area determination module 740 is used to determine human body position area to be processed according to the parameter of human body area to be processed
Domain.Specifically, automatically according to lower line coordinates, upper line coordinates, width and the central point of the human body area to be processed in image
Coordinate can determine the human body area to be processed in image.
Width detection module 750 is used to obtain the width at human body position to be processed in human body area to be processed.
In the present embodiment, the edge detection of vertical direction can be carried out to human body area to be processed automatically, obtains human body
The both sides of the edge line at position to be processed, obtain both sides of the edge line the distance between obtain the width at human body position to be processed.
Specifically, the edge detection of vertical direction is carried out to human body area to be processed, extracts the longest edge in two sides
As the both sides of the edge line of human body area to be processed, calculates the distance between both sides of the edge line and obtain human body position to be processed
Width.
The width at human body position to be processed refers to the developed width at human body position to be processed in image.
Zoom module 760 is used for each pixel in the width at human body position to be processed in human body area to be processed
The pixel value of point zooms in and out processing, the pixel value after obtaining each pixel scaling.
Specifically, scaling processing can be zoomed in and out according to scaling, or according to the zoom factor function pre-established
Zoom in and out processing.
In one embodiment, to each pixel in the width at human body position to be processed in human body area to be processed
The step of pixel value of point zooms in and out processing, obtains the pixel value after each pixel scales include:
Each pixel obtained in human body area to be processed in the width at human body position to be processed is to be processed to human body
Zoom factor corresponding to the horizontal distance of the central point of area and each pixel；
If the horizontal distance of certain pixel is less than the head width of presupposition multiple and is located at human body area to be processed
Central point the first side, then according to the central point of human body area to be processed and the corresponding zoom factor of certain pixel
First object pixel is obtained, the pixel of predetermined quantity around the first object pixel is chosen, by the picture of the predetermined quantity
The pixel value of vegetarian refreshments carries out interpolation and obtains the pixel value of interpolation, using the pixel value of the interpolation as the new of certain pixel
Pixel value；
If the horizontal distance of certain pixel is less than the head width of presupposition multiple and is located at human body area to be processed
Central point the second side, then according to the central point of human body area to be processed and the corresponding zoom factor of certain pixel
The second target pixel points are obtained, the pixel of predetermined quantity around second target pixel points is chosen, by the picture of the predetermined quantity
The pixel value of vegetarian refreshments carries out interpolation and obtains the pixel value of interpolation, using the pixel value of the interpolation as the new of certain pixel
Pixel value.
Wherein, zoom factor can be calculated according to smooth transformation formula, smooth transformation formula such as formula (1) (2) (3):
Wherein, r is the horizontal distance of the central point of pixel and human body area to be processed；D is the width on head, i.e.,
wf；A and b is constant, and the value of a and b can be determined as needed.In the present embodiment, a 1, b 0.8.Smooth transformation formula
It is according to constantly test, it is ensured that edge variation is small, and intermediate conversion is big.θ (r), λ (r) are intermediate variable, and f (r) is contracting
Put coefficient.
Each pixel (x, y) in human body area to be processed in the width at human body position to be processed, if its with
Central point (the x of human body area to be processed0, y0) horizontal distance be r, if r be less than 2d, and be located at human body portion to be processed
The first side (such as left side) of the central point in position region, then obtaining first object pixel is (x0,y0- f (r)), if r is less than 2d, and
Positioned at the second side (such as the right) of the central point of human body area to be processed, then obtaining the second target pixel points is (x0,y0+f
(r)).Wherein, the left side and the right when left and right refers to that image is checked, when personage is in facing to viewer position.
Predetermined quantity can be chosen as needed, such as four pixels, five pixels, six pixels etc..If choosing
Four pixels, then the pixel value obtained to the pixel value of four pixels using bilinear interpolation algorithm operation is as the first mesh
Mark the pixel value of pixel or the second target pixel points.Bilinear interpolation first can carry out linear interpolation in X-direction, then in the Y direction
Upper carry out linear interpolation, or linear interpolation is first carried out in the Y direction, then carries out linear interpolation in the X direction.
Zoom module 760 is also used to obtain the abscissa and ordinate of the central point of human body area to be processed, will
Abscissa of the abscissa of the central point of human body area to be processed as first object pixel, the human body is to be processed
The ordinate of the central point of area subtracts the difference of the corresponding zoom factor of certain pixel as first object pixel
Ordinate determines the first object pixel according to the abscissa of the first object pixel and ordinate；And obtain the people
The abscissa and ordinate of the central point of body area to be processed, by the horizontal seat of the central point of human body area to be processed
It is denoted as the abscissa of the second target pixel points, the ordinate of the central point of human body area to be processed is added into certain picture
Ordinate of the sum of the corresponding zoom factor of vegetarian refreshments as the second target pixel points, according to the abscissa of second target pixel points
Second target pixel points are determined with ordinate.
Above-mentioned character image processing unit, after obtaining image, face location in detection image marks the parameter of face, root
According to the parameter of face parameter estimation human body area to be processed, determine that human body waits for according to the parameter of human body area to be processed
Treatment site region obtains the width at human body position to be processed in human body area to be processed, to the picture of pixel therein
Element zooms in and out processing, and the image after being scaled realizes automatically to the pixel at the position to be processed of human body in character image
Processing, do not need to spend high cost study profession repairs figure software, automatically processes to personage in picture, high-efficient.
Fig. 8 is the structural block diagram of character image processing unit in another embodiment.As shown in figure 8, above-mentioned character image
Processing unit, in addition to include image collection module 710, detection mark module 720, estimation block 730, area determination module 740,
Width detection module 750 and Zoom module 760 further include judgment module 752 and equalization processing module 770.Wherein:
Judgment module 752 is for judging whether the width at human body position to be processed in human body area to be processed is greater than
Threshold value, if so, the Zoom module is to each pixel in the width at human body position to be processed in human body area to be processed
The pixel value of point zooms in and out processing, if it is not, then without processing.
Equalization processing module 770 is used to carry out histogram equalization processing to the image after scaling.
Specifically, the contrast of image can be improved in histogram equalization processing, and it is uneven to reduce transformation bring color.
Histogram equalization processing is to become the grey level histogram of original image in whole ashes from some gray scale interval for comparing concentration
Being uniformly distributed in degree range.
In other embodiments, a kind of character image processing unit may include image collection module 710, detection mark module
720, estimation block 730, area determination module 740, width detection module 750, Zoom module 760, judgment module 752 and equilibrium
Change any possible combination of processing module 770.
Those of ordinary skill in the art will appreciate that realizing all or part of the process in above-described embodiment method, being can be with
Relevant hardware is instructed to complete by computer program, the program can be stored in a non-volatile computer and can be read
In storage medium, the program is when being executed, it may include such as the process of the embodiment of above-mentioned each method.Wherein, the storage is situated between
Matter can be magnetic disk, CD, read-only memory (Read-Only Memory, ROM) etc..
The embodiments described above only express several embodiments of the present invention, and the description thereof is more specific and detailed, but simultaneously
Limitations on the scope of the patent of the present invention therefore cannot be interpreted as.It should be pointed out that for those of ordinary skill in the art
For, without departing from the inventive concept of the premise, various modifications and improvements can be made, these belong to guarantor of the invention
Protect range.Therefore, the scope of protection of the patent of the invention shall be subject to the appended claims.
Claims (18)
1. a kind of character image processing method, comprising the following steps:
Obtain image；
The face location in described image is detected, and marks the parameter of the face；
According to the parameter of human body area to be processed in the parameter estimation described image of the face；
Human body area to be processed is determined according to the parameter of human body area to be processed；
Obtain the width at human body position to be processed in human body area to be processed；
According to the zoom factor function pre-established, automatically to human body position to be processed in human body area to be processed
The pixel value of each pixel in width zooms in and out processing, the pixel value after obtaining each pixel scaling；Include:
Each pixel in human body area to be processed in the width at human body position to be processed is obtained to human body position to be processed
Zoom factor corresponding to the horizontal distance of the central point in region and each pixel；
If the horizontal distance of certain pixel is less than the head width of presupposition multiple and is located at human body area to be processed
First side of central point then obtains the abscissa and ordinate of the central point of human body area to be processed, by the people
Abscissa of the abscissa of the central point of body area to be processed as first object pixel, by human body portion to be processed
The ordinate of the central point in position region subtracts the difference of the corresponding zoom factor of certain described pixel as first object pixel
Ordinate determines the first object pixel according to the abscissa of the first object pixel and ordinate；And according to institute
State the new pixel value that first object pixel obtains certain pixel；
If the horizontal distance of certain pixel is less than the head width of presupposition multiple and is located at human body area to be processed
Second side of central point then obtains the abscissa and ordinate of the central point of human body area to be processed, by the people
Abscissa of the abscissa of the central point of body area to be processed as the second target pixel points, by human body portion to be processed
The ordinate of the central point in position region adds the sum of the corresponding zoom factor of certain described pixel as the second target pixel points
Ordinate determines second target pixel points according to the abscissa of second target pixel points and ordinate；And according to institute
State the new pixel value that the second target pixel points obtain certain pixel.
2. the method according to claim 1, wherein the step of parameter of the label face, includes:
Mark the center point coordinate of the face, the width of face and height；
The step of parameter of human body area to be processed, includes: in the parameter estimation described image according to the face
According to the lower line coordinates, online of the center point coordinate of the face, width and height estimation human body area to be processed
Coordinate, width and center point coordinate.
3. the method according to claim 1, wherein described obtain human body in the human body area to be processed
The step of width at position to be processed includes:
The edge detection that vertical direction is carried out to human body area to be processed, obtains the two sides at human body position to be processed
Edge line, obtain both sides of the edge line the distance between obtain the width at human body position to be processed.
4. the method according to claim 1, wherein to be processed to human body in human body area to be processed
The step of pixel value of each pixel in the width at position zooms in and out processing, obtains the pixel value after each pixel scales is also
Include:
If the horizontal distance of certain pixel is less than the head width of presupposition multiple and is located at human body area to be processed
First side of central point is then according to the central point of human body area to be processed and the corresponding scaling of certain described pixel
Number obtains first object pixel, the pixel of predetermined quantity around the first object pixel is chosen, by the predetermined number
The pixel value of the pixel of amount carries out interpolation and obtains the pixel value of interpolation, using the pixel value of the interpolation as described in certain
The new pixel value of pixel；
If the horizontal distance of certain pixel is less than the head width of presupposition multiple and is located at human body area to be processed
Second side of central point is then according to the central point of human body area to be processed and the corresponding scaling of certain described pixel
Number obtains the second target pixel points, the pixel of predetermined quantity around second target pixel points is chosen, by the predetermined number
The pixel value of the pixel of amount carries out interpolation and obtains the pixel value of interpolation, using the pixel value of the interpolation as described in certain
The new pixel value of pixel.
5. the method according to claim 1, wherein the zoom factor is calculated according to smooth transformation formula
It arrives, the smooth transformation formula are as follows:
Wherein, r is the horizontal distance of the central point of pixel and human body area to be processed, and d is the width on head, and a and b are
Constant, θ (r), λ (r) are intermediate variable, and f (r) is zoom factor.
6. the method according to claim 1, wherein obtaining people in the human body area to be processed described
After the step of width at body position to be processed, the method also includes:
Judge whether the width at human body position to be processed in human body area to be processed is greater than threshold value, if so, to institute
The pixel value for stating each pixel in human body area to be processed in the width at human body position to be processed zooms in and out processing, if
It is no, then without processing.
7. method according to any one of claim 1 to 6, which is characterized in that the method also includes:
Histogram equalization processing is carried out to the image after scaling.
8. method according to any one of claim 1 to 6, which is characterized in that the human body position to be processed is body
Or waist.
9. a kind of character image processing unit characterized by comprising
Image collection module, for obtaining image；
Mark module is detected, for detecting the face location in described image, and marks the parameter of the face；
Estimation block, the parameter for human body area to be processed in the parameter estimation described image according to the face；
Area determination module, for determining human body area to be processed according to the parameter of human body area to be processed；
Width detection module, for obtaining the width at human body position to be processed in human body area to be processed；
Zoom module, for according to the zoom factor function pre-established, automatically to people in human body area to be processed
The pixel value of each pixel in the width at body position to be processed zooms in and out processing, the pixel after obtaining each pixel scaling
Value；
The Zoom module is also used to obtain each pixel in human body area to be processed in the width at human body position to be processed
It puts to zoom factor corresponding to the horizontal distance of the central point of human body area to be processed and each pixel；
If the horizontal distance of certain pixel is less than the head width of presupposition multiple and is located at human body area to be processed
First side of central point then obtains the abscissa and ordinate of the central point of human body area to be processed, by the people
Abscissa of the abscissa of the central point of body area to be processed as first object pixel, by human body portion to be processed
The ordinate of the central point in position region subtracts the difference of the corresponding zoom factor of certain described pixel as first object pixel
Ordinate determines the first object pixel according to the abscissa of the first object pixel and ordinate；According to described
First object pixel obtains the new pixel value of certain pixel；
If the horizontal distance of certain pixel is less than the head width of presupposition multiple and is located at human body area to be processed
Second side of central point then obtains the abscissa and ordinate of the central point of human body area to be processed, by the people
Abscissa of the abscissa of the central point of body area to be processed as the second target pixel points, by human body portion to be processed
The ordinate of the central point in position region adds the sum of the corresponding zoom factor of certain described pixel as the second target pixel points
Ordinate determines second target pixel points according to the abscissa of second target pixel points and ordinate；According to described
Second target pixel points obtain the new pixel value of certain pixel.
10. device according to claim 9, which is characterized in that the detection mark module is also used to mark the face
Center point coordinate, face width and height；
The estimation block is also used to according to the center point coordinate of the face, width and height estimation human body position area to be processed
Lower line coordinates, upper line coordinates, width and the center point coordinate in domain.
11. device according to claim 9, which is characterized in that the width detection module is also used to wait for the human body
Treatment site region carries out the edge detection of vertical direction, obtains the both sides of the edge line at human body position to be processed, obtains two sides
Edge line the distance between obtain the width at human body position to be processed.
12. device according to claim 9, which is characterized in that the Zoom module is also used to:
If the horizontal distance of certain pixel is less than the head width of presupposition multiple and is located at human body area to be processed
First side of central point is then according to the central point of human body area to be processed and the corresponding scaling of certain described pixel
Number obtains first object pixel, the pixel of predetermined quantity around the first object pixel is chosen, by the predetermined number
The pixel value of the pixel of amount carries out interpolation and obtains the pixel value of interpolation, using the pixel value of the interpolation as described in certain
The new pixel value of pixel；
If the horizontal distance of certain pixel is less than the head width of presupposition multiple and is located at human body area to be processed
Second side of central point is then according to the central point of human body area to be processed and the corresponding scaling of certain described pixel
Number obtains the second target pixel points, the pixel of predetermined quantity around second target pixel points is chosen, by the predetermined number
The pixel value of the pixel of amount carries out interpolation and obtains the pixel value of interpolation, using the pixel value of the interpolation as described in certain
The new pixel value of pixel.
13. device according to claim 9, which is characterized in that
The zoom factor is calculated according to smooth transformation formula, the smooth transformation formula are as follows:
Wherein, r is the horizontal distance of the central point of pixel and human body area to be processed, and d is the width on head, and a and b are
Constant, θ (r), λ (r) are intermediate variable, and f (r) is zoom factor.
14. device according to claim 9, which is characterized in that described device further include:
Judgment module, for judging whether the width at human body position to be processed in human body area to be processed is greater than threshold
Value, if so, the Zoom module is to each picture in the width at human body position to be processed in human body area to be processed
The pixel value of vegetarian refreshments zooms in and out processing, if it is not, then without processing.
15. the device according to any one of claim 9 to 14, which is characterized in that described device further include:
Equalization processing module, for carrying out histogram equalization processing to the image after scaling.
16. the device according to any one of claim 9 to 14, which is characterized in that the human body position to be processed is body
Body or waist.
17. a kind of storage medium, is stored thereon with computer program, which is characterized in that can when described program is executed by processor
Realize such as character image processing method described in any item of the claim 1 to 8.
18. a kind of terminal device, including storage medium, processor and storage can be transported on said storage and on a processor
Capable computer program, the processor realize such as figure map described in any item of the claim 1 to 8 when executing described program
As processing method.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510613580.6A CN106558040B (en) | 2015-09-23 | 2015-09-23 | Character image treating method and apparatus |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510613580.6A CN106558040B (en) | 2015-09-23 | 2015-09-23 | Character image treating method and apparatus |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106558040A CN106558040A (en) | 2017-04-05 |
| CN106558040B true CN106558040B (en) | 2019-07-19 |
Family
ID=58413378
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510613580.6A Active CN106558040B (en) | 2015-09-23 | 2015-09-23 | Character image treating method and apparatus |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106558040B (en) |
Families Citing this family (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107395958B (en) * | 2017-06-30 | 2019-11-15 | 北京金山安全软件有限公司 | Image processing method and device, electronic equipment and storage medium |
| CN108090908B (en) * | 2017-12-07 | 2020-02-04 | 深圳云天励飞技术有限公司 | Image segmentation method, device, terminal and storage medium |
| CN110415168B (en) * | 2018-04-27 | 2022-12-02 | 武汉斗鱼网络科技有限公司 | Face local scaling processing method, storage medium, electronic device and system |
| US11410268B2 (en) | 2018-05-31 | 2022-08-09 | Beijing Sensetime Technology Development Co., Ltd | Image processing methods and apparatuses, electronic devices, and storage media |
| CN110555794B (en) * | 2018-05-31 | 2021-07-23 | 北京市商汤科技开发有限公司 | Image processing method and device, electronic equipment and storage medium |
| CN110555806B (en) * | 2018-05-31 | 2022-09-27 | 北京市商汤科技开发有限公司 | Image processing method and device, electronic device and storage medium |
| CN109034032B (en) * | 2018-07-17 | 2022-01-11 | 北京世纪好未来教育科技有限公司 | Image processing method, apparatus, device and medium |
| CN109214317B (en) * | 2018-08-22 | 2021-11-12 | 北京慕华信息科技有限公司 | Information quantity determination method and device |
| CN109472753B (en) * | 2018-10-30 | 2021-09-07 | 北京市商汤科技开发有限公司 | Image processing method and device, computer equipment and computer storage medium |
| CN111626920A (en) * | 2020-05-09 | 2020-09-04 | 北京字节跳动网络技术有限公司 | Picture processing method and device and electronic equipment |
| CN112907569B (en) * | 2021-03-24 | 2024-03-15 | 贝壳找房(北京)科技有限公司 | Head image region segmentation method, device, electronic equipment and storage medium |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6178259B1 (en) * | 1998-04-17 | 2001-01-23 | Unisys Corporation | Method of compensating for pixel histogram distortion in an object recognition system |
| CN101184143A (en) * | 2006-11-09 | 2008-05-21 | 松下电器产业株式会社 | Image processor and image processing method |
| CN101221657A (en) * | 2008-01-24 | 2008-07-16 | 杭州华三通信技术有限公司 | Image zoom processing method and device |
| CN101478629A (en) * | 2008-01-04 | 2009-07-08 | 华晶科技股份有限公司 | Image processing process for regulating human face ratio |
| CN103565441A (en) * | 2012-07-19 | 2014-02-12 | 株式会社百利达 | Biometric apparatus and computer-readable storage medium storing body image creating program |
-
2015
- 2015-09-23 CN CN201510613580.6A patent/CN106558040B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6178259B1 (en) * | 1998-04-17 | 2001-01-23 | Unisys Corporation | Method of compensating for pixel histogram distortion in an object recognition system |
| CN101184143A (en) * | 2006-11-09 | 2008-05-21 | 松下电器产业株式会社 | Image processor and image processing method |
| CN101478629A (en) * | 2008-01-04 | 2009-07-08 | 华晶科技股份有限公司 | Image processing process for regulating human face ratio |
| CN101221657A (en) * | 2008-01-24 | 2008-07-16 | 杭州华三通信技术有限公司 | Image zoom processing method and device |
| CN103565441A (en) * | 2012-07-19 | 2014-02-12 | 株式会社百利达 | Biometric apparatus and computer-readable storage medium storing body image creating program |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106558040A (en) | 2017-04-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106558040B (en) | Character image treating method and apparatus | |
| Ding et al. | Importance filtering for image retargeting | |
| US9344690B2 (en) | Image demosaicing | |
| Jia et al. | RIHOOP: Robust invisible hyperlinks in offline and online photographs | |
| US20150363922A1 (en) | Super-resolution from handheld camera | |
| CN105046708B (en) | A kind of color correction objective evaluation method consistent with subjective perception | |
| CN103745430B (en) | Rapid beautifying method of digital image | |
| CN105141838B (en) | Demosaicing methods and the device for using this method | |
| CN108419028A (en) | Image processing method, device, computer readable storage medium and electronic equipment | |
| CN107395958A (en) | Image processing method and device, electronic equipment and storage medium | |
| CN104657994B (en) | A kind of method and system that image consistency is judged based on optical flow method | |
| CN104350743B (en) | For mixed image demosaicing and the system of distortion, method and computer program product | |
| CN108961227A (en) | A kind of image quality evaluating method based on airspace and transform domain multiple features fusion | |
| Abdulrahman et al. | Color image stegananalysis using correlations between RGB channels | |
| Zhang et al. | A light dual-task neural network for haze removal | |
| CN109583341B (en) | Method and device for detecting multi-person skeleton key points of image containing portrait | |
| Rauhala et al. | A novel interface to sensor networks using handheld augmented reality | |
| CN103854020B (en) | Character recognition method and device | |
| CN110321452A (en) | A kind of image search method based on direction selection mechanism | |
| CN111275610A (en) | Method and system for processing face aging image | |
| CN106355559A (en) | Image sequence denoising method and device | |
| CN116912467A (en) | Image stitching method, device, equipment and storage medium | |
| CN113807251A (en) | Sight estimation method based on appearance | |
| CN107087114B (en) | Shooting method and device | |
| Zeng et al. | Subpixel image quality assessment syncretizing local subpixel and global pixel features |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20230625 Address after: 518057 Tencent Building, No. 1 High-tech Zone, Nanshan District, Shenzhen City, Guangdong Province, 35 floors Patentee after: TENCENT TECHNOLOGY (SHENZHEN) Co.,Ltd. Patentee after: TENCENT CLOUD COMPUTING (BEIJING) Co.,Ltd. Address before: 2, 518000, East 403 room, SEG science and Technology Park, Zhenxing Road, Shenzhen, Guangdong, Futian District Patentee before: TENCENT TECHNOLOGY (SHENZHEN) Co.,Ltd. |
|
| TR01 | Transfer of patent right |